The Restriction of the Heart Morphogenetic Field in Xenopus Levis

DEVELOPMENTAL BIOLOGY 140,328-336 (1990)

AMY K. SATER’ AND ANTONE G. JACOBSON

Department of Zoology, Center for Devebpmental Biology, University of Texas, Austin, Texa.s 78712

Accepted April 12, 1990

We have examined the spatial restriction of heart-forming potency in Xenopus 1ueG.sembryos, using an assay system in which explants or explant recombinates are cultured in hanging drops and scored for the formation of a beating heart. At the end of neurulation at stage 20, the heart morphogenetic field, i.e., the area that is capable of heart formation when cultured in isolation, includes anterior ventral and ventrolateral mesoderm. This area of developmental potency does not extend into more posterior regions. Between postneurula stage 23 and the onset of heart morphogenesis at stage 28, the heart morphogenetic field becomes spatially restricted to the anterior ventral region. The restriction of the heart morphogenetic field during postneurula stages results from a loss of developmental potency in the lateral mesoderm, rather than from ventrally directed morphogenetic movements of the lateral mesoderm. This loss of potency is not due to the inhibition of heart formation by migrating neural crest cells. During postneurula stages, tissue interactions between the lateral mesoderm and the underlying anterior endoderm support the heart-forming potency in the lateral mesoderm. The lateral mesoderm loses the ability to respond to this tissue interaction by stages 27-28. We speculate that either formation of the third pharyngeal pouch during stages 23-27 or lateral inhibition by ventral mesoderm may contribute to the spatial restriction of the heart morphogenetic field. o 1990 Academic PWSS. IW.

INTRODUCTION stages also contribute to heart formation (Ekman, 1921; Humphrey, 1972). These earlier studies included hetero- Many events of pattern formation and organogenesis topic transplantations (Copenhaver, 1924; Ekman, 1925) are mediated by inductive interactions that establish to measure the spatial extent of heart-forming potency regions of developmental potency, known as morphoge- during postneurula stages. In these experiments, the netic fields. A morphogenetic field generally comprises presumptive heart mesoderm was removed from the an- both the area of tissue that is fated to develop into a terior ventral region and replaced with grafts of meso- given structure during normal development and the tis- derm from more lateral or more posterior regions. sue immediately surrounding this area, which is capable Heart formation by the grafted tissue was interpreted of developing into the given structure under experimen- as an indicator that the grafted tissue was able to re- tal conditions (Jacobson and Sater, 1988). Extensive ex- spond to continued inductive signals. In addition, extir- perimental analysis by a number of workers has shown pation experiments measured the ability of the that the spatial extent of a morphogenetic field de- surrounding tissues to undergo regulative replacement creases with time, as the developmental potency re- in the absence of the presumptive heart mesoderm (Co- quired to make a given organ or structure is lost from penhaver, 1924). Again, cases of heart formation by the peripheral regions of the field. While the inductive in- surrounding tissues were thought to result from contin- teractions responsible for the establishment of morpho- ued induction by the underlying endoderm. Thus, earlier genetic fields have received considerable attention, the investigators often regarded tissue interactions during mechanisms underlying the spatial and temporal regula- postneurula stages as responsible for the initial estab- tion of developmental potency during subsequent devel- lishment of heart-forming potency; the induction of opment are completely unknown. heart formation during gastrula or neurula stages went We have previously shown that the inductive interac- unrecognized. tions responsible for the specification of heart meso- Tissue interactions during postneurula stages may derm occur during gastrulation in Xenopus laevis (Sater regulate the spatial extent of heart-forming potency, and Jacobson, 1989). A number of findings suggest that i.e., the heart morphogenetic field, which is induced ear- tissue interactions during neurula and postneurula lier in development. In the following experiments, we have used heart formation in explants as an assay to 1 Present address: Department of Molecular and Cell Biology, Divi- measure the spatial extent of the heart morphogenetic sion of Cell and Developmental Biology, University of California, field in postneurula-stage Xenoms embryos. These stud- Berkeley, CA 94720. ies were undertaken in order to examine the tissue in-

0012-1606/90 $3.00 328 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved. SATERANDJACOBSON Xenopus Heart Mwphogenetic Field 329 teractions governing the maintenance and restriction of 23°C). Explants and explant recombinates were cul- the heart morphogenetic field. tured in hanging drops of Niu-Twitty solution containing 50 III/ml penicillin and 50 mg/ml streptomycin MATERIALS AND METHODS (Niu-Twitty plus pen/strep) at 17”C, as described in Sater and Jacobson (1989). Grafts of embryonic tissue Embryos were allowed to heal in Niu-Twitty solution plus pen/ Embryos were obtained from both natural matings strep for up to 1 hr at 23°C; they were subsequently and in vitro fertilization. Adult Xenop were induced to maintained at 17°C. mate by a series of injections of human chorionic gonad- otropin (HCG; Sigma Chemicals, St. Louis, MO), accord- Microinjection of Lineage Tracer ing to the following schedule: males were given two in- Prior to first cleavage, embryos were transferred to jections of 150 International Units (IU) of HCG 32 and 8 5% Ficoll (Sigma) in 33% MR; embryos were then pres- hr before they were placed in the mating tank; females sure-injected with approximately 3 nl of 50 mg/ml were given one injection of 250 IU HCG 8 hr before and fluoresceinated dextran-amine (FDA) (Sigma) in dis- one injection of 500 IU HCG immediately before they tilled water, resulting in a final concentration of approx- were placed in the mating tank. Frogs were allowed to imately 150 rig/embryo. Micropipets with an external mate overnight. The following morning, embryos were tip diameter of 20-25 pm were prepared by pulling boro- collected and dejellied for 5 min in 2% cysteine HCl, pH silicate glass microcapillary tubes on a Brown-Flaming 7.4. Dejellied embryos were washed extensively in 10% electrode puller (Sutter Instruments, San Francisco, Holtfreter’s solution (Jacobson, 1966) and maintained in CA). Micropipets were calibrated by expelling FDA into 10% Holtfreter’s solution at 1’7°C. All embryos were a drop of mineral oil and measuring the diameter of the staged according to Nieuwkoop and Faber (1967). expelled FDA. Following microinjection, embryos were In vitro fertilizations were carried out by the methods allowed to heal in 33% MR containing 5% Ficoll. Em- of Gimlich and Gerhart (1984). Ovulation was induced bryos that leaked cytoplasm were discarded. by an injection of 500 IU HCG (Sigma). Approximately 12 hr later, oocytes were stripped into a petri dish con- Histology taining minced testis tissue in a minimal volume of 33% modified amphibian Ringer (MR; 100% MR: 0.1 MNaCl, Embryos containing grafts of FDA-labeled tissue 2 mM KCl, 2 mM CaCl,-2H,O, 1 mM MgCl,-GH,O, buff- were fixed in freshly prepared 2% paraformaldehyde in ered to pH 7.4 with NaHCO,) (Gimlich and Gerhart, 0.1 M Na cacodylate, pH 7.4, for approximately 20 hr at 1984). Oocytes and sperm were mixed gently by rotating 6°C. Embryos were then washed 15-24 hr in 0.1 M Na the dish and then allowed to stand for several minutes cacodylate, pH 7.4, at 6°C. Embryos were dehydrated so that fertilization could occur. Oocytes were then de- through an ethanol-butanol series, embedded in Para- jellied as described earlier and maintained in 33% MR plast, and sectioned at 8 pm. Sections were collected on at 17°C. gelatin-coated slides, dewaxed through a xylene-ethanol series, and mounted in 80% glycerol containing 4% N-propyl gallate (Sigma; Gimlich and Braun, 1985). Microsurgery and Culture of Explants and Sections were viewed using epifluorescence optics Explant Recombinates (Zeiss). For microsurgery, embryos were transferred to petri dishes containing sterile, freshly prepared Niu-Twitty RESULTS solution over a cushion of 2% agar. All embryonic oper- The Heart Mvhogenetic Field at the End ations were performed using electrolytically sharpened of Neurulation tungsten needles, sharpened watchmaker’s forceps, and eyebrow hair knives. Operations involving the separa- At the end of neurulation (stage 20), the two regions tion of mesoderm and endoderm were performed in Niu- of prospective heart mesoderm, which originally lay at Twitty solution containing 0.01% trypsin (Sigma Type the anterior lateral edge of the mesodermal mantle, IX), as described by Slack (1984). Explants isolated in have migrated ventrally to fuse at the ventral midline in the presence of trypsin were subsequently incubated in the anterior region (Nieuwkoop and Faber, 1967). Thus, 0.02% soybean trypsin inhibitor (SBTI; Sigma Type II- by stage 20, the prospective heart mesoderm has arrived S) in Niu-Twitty solution for several minutes before at the site of heart morphogenesis and undergoes no they were placed in culture. Explant recombinates were further large-scale movements. A crude map of the allowed to heal in Niu-Twitty solution containing SBTI heart morphogenetic field at this stage was constructed for up to 1 hr at room temperature (approximately by culturing ventral and lateral explants from anterior, DEVELOPMENTALBIOLOGY VOLUME 140,1990

vesicles, which were examined with a dissecting microscope. Within 3 to 4 days after explantation, explants became inflated and translucent; explants were subsequently examined with a compound microscope and transmitted light. Explants isolated from ventral regions always exhibited large, well-developed hearts, which appeared within 5 days after explantation. In contrast, the timing of heart formation in explants of lateral tissues varied considerably. Hearts formed from explants of lateral tissues exhibited much greater midanterior midposterior range of variation in the degree of heart development, from narrow strands or knots of cells contracting in a coordinated “heartbeat” pattern to well-developed hearts with the typical “looping and bending” morphol- ogy that delineates the different regions of the heart. Foci of beating tissue were easily distinguishable from strands of skeletal muscle that were occasionally observed in larger explants, since the former exhibited a

oosrerior regular peristaltic or metachronal beat pattern, in contrast to the erratic simple contractions of skeletal mus-

0 cle fibers. Generally, hearts developing from lateral explants were smaller than those developing from ventral 11 explants. Other organs or tissues observed in these explant cultures included mesenchyme and mesentery, which were 0 present in nearly all explants, and the cement gland,

7 which appeared only in ventral explants. Lateral explants often contained pronephric tubules, visible as FIG. 1. The heart morphogenetic field at the end of neurulation coiled tubes containing beating cilia. Pronephric tubules (stage 20). Lateral and ventral explants were isolated from four regions along the anterior-posterior axis. Beating hearts formed only in were not observed in ventral explants. lateral and ventral explants from the anterior region. Spatial Restrictim of the Heart Morphogenetic Field during Postneurula Stages midanterior, midposterior, and posterior regions of stage 20 embryos and scoring for heart formation; these Changes in the spatial distribution of heart-forming explants included mesoderm, endoderm, and epidermis. potency during postneurula stages were investigated by A diagram of the experiment and a summary of the re- isolating anterior lateral and anterior ventral regions at sults are shown in Fig. 1. Hearts formed only in anterior successive stages between the end of neurulation at ventral and anterior lateral explants; no hearts were stage 20 and the onset of heart morphogenesis at stage observed in explants removed from more posterior re- 28. Each explant contained mesoderm, endoderm, and gions. Explants of anterior ventral tissue, which in- epidermis, Explants were maintained in hanging drop cluded the mesoderm fated to give rise to the heart, cultures and scored for heart formation. formed beating hearts in 100% of cases, while explants The frequency of beating heart formation in lateral of anterior lateral tissue formed beating hearts in 69% and ventral explants from postneurula stages is shown of cases. These results demonstrate that at the end of in Fig. 2. With the exception of a single case, explants of neurulation, the heart morphogenetic field includes ventral regions always formed beating hearts. both anterior lateral and anterior ventral mesoderm. During early postneurula stages (stages 20 to 22), the frequency of heart formation in lateral explants is high, starting at 69% at stage 20 and increasing to 83% by Description of D$wentiation and Development stage 22. After stage 22, however, the frequency of heart of Explan.ts formation in lateral explants begins to decrease. In lat- Explants were first examined 2 to 3 days after eral explants removed at stage 23, the frequency of explantation and daily or every other day thereafter for heart formation is approximately equal to that observed at least 2 weeks. Initially, explants healed into opaque in lateral explants removed at stage 20. The frequency of SATERANDJACOBSON Xenopus Heart Morphogenetic Field 331

120 -

100 I p...*- ...... 1”““...... _.. l...... --*. .V” .. . . . *” ...... * ...... 1 ...... *- ventralmesodcnn + endoderm -;- lareralmesodenn + endodcrm l - - lateral mesoderm only

time (stages)

FIG. 2. The frequency of heart formation in mesodermal explants from postneurula stages. Sample sizes for explants of ventral mesoderm plus endoderm: st. 20, n = 24; st. 21, n = 19; st. 22, n = 13; st. 23, n = 24; st. 25, n = 26; st. 26, n = 24; st. 27, n = 8; st. 28, n = 8. Sample sizes for explants of lateral mesoderm plus endoderm: st. 20, n = 29; st. 21, n = 26; st. 22, n = 23; st. 23, n = 24; st. 24, n = 20; st. 25, n = 29; st. 26, n = 24; st. 27, n = 16; st. 28, n = 15. Sample sizes for explants of lateral mesoderm only: st. 20, n = 22; st. 21, n = 28 st. 22, n = 24; st. 23, n = 32; st. 24, n = 31; st. 25, n = 20; st. 26, n = 14. heart formation in lateral explants removed at late Approximately 50% of the transplants healed properly, postneurula stages is very low: explants removed at and only these were analyzed. Operated embryos were stage 2’7 form beating hearts in only 12.5% of cases. allowed to heal in Niu-Twitty solution and fixed at Lateral explants removed from embryos at stage 28 stage 28. never form beating hearts. A section from a transplant containing a labeled graft is shown in Fig. 3. The labeled mesoderm remains in the lateral region; none of the labeled mesoderm appears in Restriction of the Heart Morphogenetic Field Is Not Due the region surrounding the ventral midline. This region, to Cell Migration containing the presumptive heart mesoderm, lies imme- The preceding experiment demonstrates that the heart morphogenetic field becomes spatially restricted to the ventral mesoderm during postneurula stages. Two hypotheses may account for this. First, the lateral mesoderm may lose its ability to participate in heart formation. Second, the lateral mesoderm may migrate ventrally during postneurula stages; in this case, the restriction of the heart morphogenetic field results f rom morphogenetic movements, rather than from a 11ass of developmental potency. This latter possibility was tested by grafting anterior 1,ateral mesoderm labeled with the lineage tracer FDA nto the anterior lateral regions of unlabeled host em- kbryos at stage 20. These embryos were fixed at stage 28, sectioned, and examined by fluorescence microscopy. Labeled embryos to serve as transplant donors were prepared by injecting embryos prior to first cleavage with FDA, as described above. At stage 20, anterior lateral mesoderm plus the overlying epidermis was removed from the labeled embryo and grafted into the anterior lateral region of an unlabeled host. The operations were performed so that the transplanted tissue represented as nearly as possible the area included in explants of anterior lateral mesoderm; the ventral mar- FIG. 3. Section through embryo containing a graft of anterior lat- gin of each graft was placed immediately lateral and eral mesoderm labeled with FDA. Embryo is at stage 28. N, notochord; posterior to the lateral margin of the host cement gland. V, ventral midline. 332 DEVELOPMENTALBIOLOGY VOLUME 140,199O diately posterior to the cement gland, and explants of TABLE 1 this ventral region are as wide as the cement gland. At FREQUENCYOF HEART FORMATIONIN EXPLANTS FROMEMBRYOS LACKING THE NEURAL CREST stage 20, the cement gland is considerably wider than it is at stage 28; this narrowing is probably due to changes Stage at explant: 23 24 25 26 27 in the shape of the cement gland cells and cell move- Explants ments related to the anterior-posterior elongation occurring during these stages. Thus, explants of ventral Ventral mesoderm + endoderm regions taken from stage 20 embryos are wider than No. forming hearts 16 9 9 12 5 ventral region explants taken from stage 28 embryos. Total no. of cases 16 9 9 12 5 Since the labeled mesoderm does not enter the ventral % forming hearts 100 100 190 loo 100 region, one can exclude the possibility that ventralward % heart formation in migration of lateral mesoderm is principally responsi- explants from intact embryos 100 95 100 100 100 ble for the apparent restriction of the heart morphogenetic field. Moreover, in transplant embryos allowed to Lateral mesoderm develop until stage 34, none of the labeled graft cells f endoderm No. forming hearts 15 6 7 4 1 appear in the heart itself (data not shown). Therefore, Total no. of cases 26 13 18 21 10 the spatial restriction of the heart morphogenetic field % forming hearts 58 46 39 19 10 is not due to morphogenetic movements, but rather to a % heart formation in loss of heart-forming potency from lateral mesoderm. explants from intact embryos 71 50 34 46 12.5 Interactions between Mesoderm and Endoderm during Postneurula Stages Note. The neural crest and neural tube were removed from donor embryos at stage 20-21. To determine whether interactions between the lateral mesoderm and the underlying anterior lateral endoderm are involved in the maintenance of heart-form- Duncan, 1968). Therefore, neural crest cells might sup- ing potency in lateral mesoderm during postneurula press heart-forming potency in the lateral mesoderm stages, explants of lateral mesoderm were removed while migrating through it. To test this hypothesis, the from postneurula stage embryos and cultured in the ab- neural crest and neural tube were removed from em- sence of anterior endoderm. The frequency of heart for- bryos at stage 20, when the neural crest itself had mation in explants of lateral mesoderm from postneu- formed above the neural tube along its entire length rula embryos is shown in Fig. 2. Explants of lateral me- (Nieuwkoop and Faber, 1967). Neural crest migration soderm removed from embryos at stage 20 undergo progresses in a cranial-to-caudal fashion during stages heart formation in 36% of cases. For explants removed 19 to 31 (Nieuwkoop and Faber, 1967). Because cranial at stage 23, the frequency of heart formation increases neural crest cells begin migrating almost immediately to 66%, dropping to 14% for explants removed at stage after the appearance of the neural crest itself, the 26. At early postneurula stages (stages 20 to 22), the neural tube was removed in addition to the crest, in frequency of heart formation in explants of lateral me- order to remove any neural crest cells that had begun to soderm is considerably lower than the frequency of move out along the surface of the neural tube. After heart formation in lateral explants containing both me- surgery, the embryos were cultured until intact controls soderm and endoderm. This result indicates that inter- had reached mid- to late postneurula stages (stages 23 actions between the lateral mesoderm and the endo- to 26). Lateral and ventral explants of embryos sub- derm support heart-forming potency during these jected to the operation were then prepared as described stages. However, during subsequent stages, the fre- earlier; each explant contained mesoderm, endoderm, quency of heart formation in explants of lateral meso- and epidermis. These explants were scored for the fre- derm plus endoderm is only slightly higher than that quency of beating heart formation. observed in explants of lateral mesoderm alone, sug- If migrating neural crest cells inhibit heart-forming gesting that this supportive interaction is not effective potency in the lateral mesoderm, then removal of the after stage 23. neural crest should result in an increase in the fre- POSSIBLE CAUSES OF THE RESTRICTION OF THE quency of heart formation in lateral explants at later HEART MORPHOGENETIC FIELD stages, relative to the frequency observed in lateral ex- Restriction of the Heart Morphogenetic Field Is Not Due plants from unoperated embryos. As the results shown to Local Suppression of Heart Formation by in Table 1 indicate, this prediction is not born out: at Migrating Neural Crest Cells nearly all stages tested, the frequency of heart forma- Neural tissues are known to inhibit heart formation tion in lateral explants from operated embryos is lower during neurula stages in urodele embryos (Jacobson and than that observed in lateral explants from intact em- SATERAND JACOBSON Xenopus Heart Morphogenetic Field 333

TABLE 2 combinates formed beating hearts, indicating that the FREQUENCY OF HEART FORMATION IN HETEROCHRONIC EXPLANT operation itself does not reduce heart-forming potency. RECOMBINATIONS OF LATERAL MESODERM AND LATERAL ENDODERM These results demonstrate that the lateral mesoderm No. of % of loses the competence to respond to signals from the ante- Stage of Stage of recombinates recombinates rior endoderm during late postneurula stages; this loss mesoderm endoderm forming Total no. of forming may contribute to the spatial restriction of the heart at explant at explant hearts recombinates hearts morphogenetic field.

27-28 20-23 1 14 7 20 20 8 9 89 DISCUSSION Our results lead to three major conclusions. First, the heart morphogenetic field becomes restricted to the anterior ventral mesoderm during postneurula stages. Sec- bryos. Thus, inhibition of heart formation by migrating ond, this spatial restriction is due to a change in the neural crest cells does not account for the restriction of developmental potency of the lateral mesoderm. Third, the heart morphogenetic field. heart-forming potency in peripheral regions of the morphogenetic field is supported by the underlying endo- Loss of Responsiveness in Lateral Mesoderm Contributes derm. to the Restriction of the Heart Morphogenetic Field The Heart Morphogenetic Field Becomes Spatially Since interactions between the anterior endoderm Restricted following Neurulation and the anterior lateral mesoderm serve to maintain the heart morphogenetic field in the lateral mesoderm dur- A specification map of the lateroventral regions of the ing early postneurula stages (stages 20 to 23), the per- Xenoms embryo at stage 20 indicates that explants of turbation of these interactions may bring about the anterior lateral and anterior ventral regions will un- subsequent restriction of the heart morphogenetic field. dergo heart formation in culture. By the onset of heart These interactions would be disrupted if either the ante- morphogenesis at stage 28, this heart-forming capacity rior endoderm failed to produce the appropriate signal is restricted to the anterior ventral region. At early after early postneurula stages, or the anterior lateral postneurula stages, it is unclear whether heart-forming mesoderm lost the ability to respond to signals produced potency is retained by all cells throughout the lateral by the anterior endoderm. Experiments to determine region. It seems more likely that heart formation in lat- whether the endoderm continues to produce the appro- eral explants results from the retention of heart-form- priate signal are infeasible because of the difficulties of ing potency by cells at or near the ventral edge of the designing an appropriate assay. In contrast, it is rela- lateral region. Cells in more dorsal areas of the lateral tively easy to determine whether the responsiveness of region may not participate in heart formation in these the lateral mesoderm persists into late postneurula explants, instead contributing to the formation of mes- stages by examining the interactions between late post- enchyme, mesothelium, or pronephros. neurula lateral mesoderm and early postneurula endo- Our description of the heart morphogenetic field dur- derm. ing postneurula stages is in general agreement with the Heterochronic explant recombinations were prepared results of heterotopic transplantation and tissue re- to determine whether interactions with endoderm from moval experiments by Copenhaver (1924) and Ekman early postneurula embryos could restore heart-forming (1925), which show that the distribution of heart-form- potency in late postneurula lateral mesoderm. Explants ing potency is limited to anterior ventral and ventrola- of anterior lateral mesoderm from late postneurula em- teral mesoderm in amphibian embryos. The concur- bryos (stages 2’7 to 28) were recombined with anterior rence of findings on the heart morphogenetic field of lateral endoderm from embryos at stages 20 to 23. As a amphibians is in marked contrast to studies of the heart control, explant recombinates were prepared in which morphogenetic field in chick embryos, in which differ- both the anterior lateral endoderm and the anterior lat- ent experimental approaches have yielded conflicting eral mesoderm were isolated at stage 20. Recombinates results. Heart-forming potency is expressed by explants were cultured in hanging drops and scored for the fre- removed from large areas of the peripheral blastoderm quency of heart formation. The results are shown in of chick embryos at the regressing primitive streak Table 2. Out of 14 successful cases, only one hetero- stage (Rawles, 1936). However, at the same stage, re- chronic recombinate formed a beating heart, indicating moval of the prospective heart mesoderm from one side that interaction with early postneurula endoderm does of the embryo prevents the expression of heart-forming not restore heart-forming potency in late postneurula potency by any tissue on the operated side (DeHaan, lateral mesoderm. In contrast, 89% of the control re- 1965). Heart formation does not occur in chick embryos 334 DEVELOPMENTALBIOLOGY vOLUME140.1900 from which both regions of prospective heart mesoderm represent an induced morphogenetic field, capable of have been removed. heart formation even in the absence of continued induc- These differences underscore the importance of the tive signals. To distinguish between these possibilities, choice of experimental manipulation used to define the we cultured explants of anterior lateral mesoderm both spatial extent of a morphogenetic field. Observation of with and without the underlying endoderm. Our results the developmental potency expressed by explanted tis- indicate that anterior lateral endoderm supports heart- sues that are cultured in isolation, operationally defined forming potency in anterior lateral mesoderm during as the “state of specification” of the explanted tissue early postneurula stages: explants containing both me- (Slack, 1983), may be the most easily interpretable way soderm and endoderm undergo heart formation more to examine the spatial extent of a morphogenetic field. frequently than do explants containing mesoderm Such observations reflect only those tissue interactions alone. The specificity of the interaction between the lat- that occurred prior to the removal of the tissue from the eral mesoderm and the underlying endoderm has not embryo; further inductive or inhibitory signals that been examined. Our results suggest, however, that the might alter the expression of developmental potency fol- ability to respond to continued signals from the anterior lowing a heterotopic graft or an extirpation experiment endoderm may be a critical characteristic of mesoderm are avoided. However, in amphibians, different experi- in peripheral regions of the heart morphogenetic field. mental manipulations have produced similar maps of Two lines of evidence suggest that this supportive in- the heart morphogenetic field. This common result teraction does not persist during late postneurula lends support to the idea of the heart morphogenetic stages. First, the difference in the frequency of heart field as a “dynamic region of developmental potency” formation between late postneurula mesodermal ex- (Jacobson and Sater, 1988), including both the anterior plants with and without endoderm is statistically insig- ventral mesoderm, fated to become the heart, and the nificant. In addition, anterior lateral endoderm from adjacent lateral mesoderm, fated to contribute to the early postneurula stages, which supports heart-forming vascular and skeletal elements of the branchial arches. potency in early postneurula lateral mesoderm, does not Moreover, it suggests that inhibitory signals arising in restore heart-forming potency in late postneurula lat- tissues outside the morphogenetic field do not delimit or eral mesoderm. This finding suggests that by late post- restrict the heart morphogenetic field. If the heart mor- neurula stages, the lateral mesoderm has lost its ability phogenetic field were restricted by inhibitory signals to respond to anterior endoderm. The loss of responsive arising outside the field itself, one would expect the re- ability during late postneurula stages may contribute to gion of heart-forming potency on a map of the heart the decline in heart-forming potency in the lateral me- morphogenetic field produced by grafting experiments soderm, and thus to the spatial restriction of the heart to be smaller than that observed on a map produced by morphogenetic field. explant experiments, since only the grafted tissues Similar supportive signals, traditionally referred to would be subjected to such signals. as “formative influences” (Holtfreter and Hamburger, Our previous work has shown that the inductive in- 1955), may be the primary target of the cardiac lethal teractions responsible for the specification of heart me- mutation in axolotls. This mutation is characterized by soderm are complete well before the end of neurulation the formation of a heart tube that fails to beat; heart at stage 20 (Sater and Jacobson, 1989). At this stage, the mesoderm will, however, develop normally when placed heart morphogenetic field comprises lateral mesoderm in a wild-type host (Humphrey, 1972). The mutation as well as the ventral mesoderm fated to give rise to the causes a defect in the endoderm that prevents it from heart. The work of Fullilove (1970) on heart-inducing supporting the later stages of heart differentiation (Le- tissues in urodele neurulae shows that the areas capable manski et al., 1979). The spatial pattern of supportive of inducing heart formation include anterior dorsolat- interactions may help to delimit morphogenetic fields in era1 and ventral endoderm, a substantial portion of the a variety of instances. total endoderm. Thus, during the establishment of heart mesoderm, the regions of inductive capacity and heart- The Restriction of the Heart Morphogenetic Field forming potency are considerably larger than the re- Represents a Change in Developmental Potency gions destined to participate in heart formation. Analyses of changes in the spatial distribution of any Interactions between Lateral Mesoderm and Lateral field over time must examine the possibility that mor- Endoderm in Postneurula Embryos phogenetic movements may be responsible for such Ekman (1921,1925) viewed the region of heart-form- changes. Therefore, spatial restriction of the morphoge- ing potency described by his heterotopic transplanta- netic field may reflect cell movements, rather than a tion and tissue removal experiments as a region that restriction of developmental potency. Our results indi- possesses the competence to respond to the signals in- cate that the decline in heart-forming potency in ex- ducing heart formation. Alternatively, this region may plants of anterior lateral mesoderm during postneurula SATERANDJACOBSON Xenopus Heart Morphogenetic Field 335 stages is not due to ventralward migration of lateral ward the ectoderm, moving through the intervening mesodermal cells that retain heart-forming potency, mesoderm. A corresponding invagination of the pharyn- because no such migration occurs. Thus, the spatial re- geal ectoderm occurs at the same time. Eventually, the striction of the heart morphogenetic field is caused by a endoderm at the tip of the pharyngeal pouchcomes into loss of heart-forming potency within the lateral meso- contact and fuses with the invaginating ectoderm. Six derm itself. pharyngeal pouches form in a cranial-to-caudal se- quence in vertebrate embryos, dividing the anterior Possible Causes of the Restriction of the Heart ventrolateral mesoderm into separate areas. Morphogenetic Field The third pharyngeal pouches first appear as a dim- Our results also exclude the possibility that neural pling in the ventrolateral pharyngeal endoderm lateral crest cells suppress heart-forming potency in the lateral and slightly dorsal to the prospective heart mesoderm mesoderm during their migration into the branchial re- at stage 23 (Nieuwkoop and Faber, 1967). The evaginat- gion, since embryos lacking the neural crest and neural ing endoderm of the third pouch makes initial contact tube do not exhibit greater heart-forming potency with the ectoderm at stage 27 (Nieuwkoop and Faber, within the lateral mesoderm than do intact embryos. 1967). Thus, formation of the third pharyngeal pouch Thus, restriction of the heart morphogenetic field is not occurs during the period when the anterior lateral me- due to a local inhibition of heart-forming potency by soderm is losing heart-forming potency. In addition, the migrating neural crest cells. This is in contrast to the third pharyngeal pouch divides the anterior lateral me- findings of Jacobson and Duncan (1968), who reported soderm, perhaps preventing communication between that urodele neural tissue, including the neural plate the regions of lateral mesoderm anterior and posterior and, later, the neural crest, inhibited heart formation to the pouch. The observation that formation of the when combined with explants of prospective heart me- third pharyngeal pouch coincides spatially and tempo- soderm in culture. In addition, it has more recently been rally with the loss of heart-forming potency in the ante- suggested that neural crest cells participate in the re- rior lateral mesoderm suggests that pouch formation striction of the lens morphogenetic field by locally in- may contribute to the loss of heart-forming potency. hibiting lens-forming potency in the epidermis (Henry Specifically, the third pharyngeal pouch may divide the and Grainger, 1987). The differences between the re- anterior lateral mesoderm into isolated regions of tissue sults of Jacobson and Duncan and those reported here that are too small to sustain heart-forming potency, may reflect either differences in the tissue interactions leaving them susceptible to incorporation into the bran- of urodele and anuran embryos, or changes in the sensi- chial arch skeleton and vascular elements. tivity of the heart mesoderm to inhibitory tissue inter- The loss of developmental potency as a result of the actions over the course of development. subdivision of a morphogenetic field beyond a critical The restriction of the heart morphogenetic field may point has substantial precedent. Holtfreter and Ham- result from interactions within the field itself. During burger (1955) described a number of studies demonstrat- postneurula stages, the anterior ventral mesoderm, ing that the range of differentiated cell types exhibited which is fated to give rise to the heart, may act to sup- by an isolated region of developmental potency is press heart-forming potency in the anterior lateral me- greatly reduced when the isolated tissue is divided into soderm. This type of local inhibition is analogous to the fragments which are then cultured individually. This suppression of bristle formation within a small radius phenomenon has been exhaustively studied by Ban- of a bristle organ primordium in the epidermis of Dro- Holtfreter (1965), who examined the differentiative ca- sophila and other insects (Stern, 1954). One model sug- pability of isolated regions of the dorsal lip of the blas- gests that short-range inhibition of bristle formation by topore in urodele embryos. She found that explants of the newly committed bristle organ primordium results the entire dorsal lip exhibited a wide range of differen- in the ordered spatial array of bristles that is character- tiated cell types, including cell types to which the dorsal istic of insect epidermis (Wigglesworth, 1940; Held and lip does not contribute in viva. Smaller fractions of the Bryant, 1984). Such tissue interactions serve to organize dorsal lip produced a much more limited diversity of developmental processes within morphogenetic fields differentiated tissues. Differentiation in the smallest and may play a large and as yet unrecognized role in the pieces produced very few cell types: in some of these assignment of cell fate in embryos. explants, recognizable differentiation did not occur. The A more proximate cause of the loss of heart-forming smaller explants were unable to give rise to the full potency from the anterior lateral mesoderm during range of cell types which the explanted regions would postneurula stages may rest in the formation of the generate in vivo. The limited differentiative capability third pharyngeal pouch. During pharyngeal pouch for- of very small explants has been attributed to “mass ef- mation, the endoderm lining the lateral walls of the fects”: because of small cell number, cell-cell interac- archenteron in the pharyngeal region evaginates to- tions occur at levels insufficient to support a wide range 336 DEVELOPMENTALBIOLOGY VOLUME 140,199O of differentiative capabilities (Muchmore, 1951). An al- DEHAAN, R. L. (1965). Morphogenesis of the vertebrate heart. In “Or- ternative explanation is that very small explants are ganogenesis” (R. L. DeHaan and H. Ursprung, Eds.), pp. 377-419. incapable of “conditioning” the culture medium with se- Holt, Rinehart, and Winston, New York. creted products that are necessary to support the for- EKMAN, G. (1921). Experimentelle Beitrage zur Entwicklung des Bom- binatorherzens. Oversikt. av. Finska Vetenskapssoeietetens For- mation of diverse cell types (Niu and Twitty, 1953). Re- handlingar 63,1-37. cently Gurdon (1988,1989) has reinvestigated these phe- EKMAN, G. (1925). Experimentelle Beitrage zur Herzentwicklung der nomena, showing that Xenopus animal cap cells are Amphibien. Wilhelm Rcn~x’Arch. Entwicklungsmech Org. 106,320- unable to express cardiac actin in response to mesoder- 352. ma1 induction unless they are partially surrounded by FULLILOVE, S. L. (1970). Heart induction: Distribution of active factors in newt endoderm. J. Exp. Zoo1 175.323-326. other cells that are also responding to the inductive sig- GIMLICH, R. L., and BRAUN,J. (1985). Improved fluorescent compounds nal. He refers to this requirement for similarly re- for tracing cell lineage. Dev. Biol. 28,454-4’71. sponding cells as a “community effect.” GIMLICH, R. L., and GERHART,J. C. (1984). Early cellular interactions It is clear that events occurring within the lateral me- promote embryonic axis formation in Xenopus Levis. Dev. Biol. 104, 117-130. soderm contribute to the loss of heart-forming potency GURDON,J. B. (1988). A community effect in animal development. Na- in this region. By late postneurula stages, the lateral ture (I&n&m) 336,772-774. mesoderm has lost the ability to participate in tissue GURDON,J. B. (1989). The localization of an inductive response. Devel- interactions with the anterior lateral endoderm that opment 105,27-33. support heart-forming potency of the lateral mesoderm. HAMBURGER,V. (1960). A Manual of Experimental Embryology, re- vised ed. University of Chicago Press, Chicago, IL. It is considerably less clear, however, which cellular HELD, L. I., and BRYANT, P. J. (1984). Cell interactions controlling the processes govern this loss of developmental potency. formation of bristle patterns in Drosophila. In “Pattern Forma- Numerous experimental analyses of the role of induction” (G. M. Malacinski and S. V. Bryant, Eds.). MacMillan Publish- tive interactions in embryonic determination indicate ing Co., New York. that inductive interactions occur gradually, with a cu- HENRY, J. J., and GRAINGER,R. M. (1987). Inductive interactions in the mulative effect on the developmental potency of the re- spatial and temporal restriction of lens-forming potential in embryonic ectoderm of Xenopus laevis. Dev. Biol 124,200-214. sponding tissue (Jacobson, 1966). It follows that the ac- HOLTFRETER,J., and HAMBURGER,V. (1955). Amphibians. In “Analy- quisition of developmental potency also occurs gradu- sis of Development,” pp. 230-296. Saunders, Philadelphia. ally, and that the induced state of determination is HUMPHREY, R. R. (1972). Genetic and experimental studies on a mu- labile for some period of time prior to a final, relatively tant gene (c) determining absence of heart action in embryos of the Mexican Axolotl (Ambystoma mexicanurn). Dev. Biol. 27,365-375. irreversible commitment to a given developmental JACOBSON,A. G. (1966). Inductive processes in embryonic develop- pathway. Perhaps the spatial restriction of the heart ment. Science 152,25-34. morphogenetic field and the attendant loss of develop- JACOBSON,A. G., and DUNCAN, J. T. (1968). Heart induction in sala- mental potency from peripheral regions of the heart manders. J. Exp. Zool. 167,79-103. JACOBSON,A. G., and SATER, A. K. (1988). Features of embryonic in- morphogenetic field represent a failure of an induced duction. Development 104,341-359. state of determination to become “fixed” within the MUCHMORE,W. B. (1951). Differentiation of the trunk mesoderm in cells at the periphery of the morphogenetic field. The Amblvstoma madatum J. Exp. Zool. 118,137-186.--_. retention of developmental potency by cells that become NIEUW~OOP,P. D., and FABER, J. (1967). “Normal Table of Xenopus committed to a developmental pathway, on the other Levis (Daudin).” North-Holland, Amsterdam. LEMANSKI, L. F.,‘PAULSON,D. J., and HILL, C. S. (1979). Normal ante- hand, may result from cellular processes that stabilize rior endoderm corrects the heart defect in cardiac mutant sala- the induced state of determination. manders (Ambystoma mexicanum). Science 204,860-862. NIU, M. C., and TWITS, V. C. (1953). The differentiation of gastrula We thank Dr. Akif Uzman for extensive discussion and review of ectoderm in medium conditioned by axial mesoderm. Proc. Nat. the manuscript, and Jon Nuelle for insightful comments and assis- Acad Sci. USA 39,985-989. tance with the microinjection of lineage tracer. This work was sup- RA~LES, M. E. (1936). A study in the localization of organ-forming ported by grants from the American Heart Association, Texas Affili- areas in the chick blastoderm of the head process stage. J. Exp. Zool. ate, and the University of Texas University Research Institute. A.K.S. 72,271-315. was supported by graduate fellowships from the National Science SATER, A. K., and JACOBSON,A. G. (1989). The specification of heart Foundation and the University of Texas. mesoderm occurs during gastrulation in Xenqwus laevis. Develop ment 105,821-830. STERN,C. (1954). Genes and developmental patterns. Proc. Ninth ht. REFERENCES Chg. Gen. 1,355-369. SLACK, J. M. W. (1983). “From Egg to Embryo. Determinative Events BAN-H• LTFRETER, H. (1965). “Differentiation Capacities of Spe- in Early Development.” Cambridge Univ. Press, Cambridge. mann’s Organizer Investigated in Explants of Diminishing Size.” SLACK, J. M. W. (1984). Regional biosynthetic markers in the early Ph.D. dissertation, University of Rochester. University Microfilms, amphibian embryo. J. Embryol. Exp. Morph01 80,289-319. Ann Arbor. WIGGLESWORTH,V. B. (1940). Local and general factors in the develop- COPENHAVER,W. M. (1924). Experiments on the development of the ment of “pattern” in Rhodnius prolixus (Hemiptera). J. Exp. Zool heart of Amblystoma punctatum. J. Exp. ZooL 43,321-371. 17,180-200.